Development And Research Of Abnormal Vibration Intelligent Diagnosis Technology For Rolling Mill

Steel is the essential material for our country’s constructing, and strip hot rolling mill is a primary manufacturing equipment to produce steel. The producing line of Strip hot rolling mill had achieved good economic benefits when it was put into operation. However, with the improvement of manufacturing equipment and production process, increase in rolling speed, and lower in rolling thickness, the process conditions became complicated, the strip hot rolling mill had a variety of vibration modes and a lot of new fault, brought adverse effects to the steel producing. Research and develop a intelligent fault diagnosis system is necessary to preclude fault and protect the producing line, the system not only can bring economic benefits to the community, but also have important significance in theory.First of all, this paper introduces internal and foreign state of strip hot rolling mill based on the Zhujiang Steel Co.,Ltd strip hot rolling mill, and introduce the source and research significance of the subject. Then research the type of the strip hot rolling mill system vibration and the mechanism of strip hot rolling mill, analyze the effect of rolling parameters for strip hot rolling mill in rolling process, it is helpful to research and develop the fault diagnosis expert system.According to the actual situation of Zhujiang strip hot rolling mill, determine the vibration parameters, make a vibration test plan and selected the test equipments. Analysis the collected dates with some methods such as time domain, cross-spectral, cross-correlation, had some relevant conclusions, find out the vibration reasons of strip hot rolling mill, and give some solutions from mechanical design, process parameters and device status.In fault diagnosis applications, neural network and expert system are two important technology in artificial intelligence field. The paper establish neutral network model, research the algorithm of back propagation and give improved method. Finally, discuss the machine learning and reason the mechanism of expert system.On the basis of user needs, finished the strip hot rolling mill’s Abnormal vibration intelligent diagnosis expert system which base on the neutral network by using VB.NET as the language of system development. After the system’s being tested and verified, it achieved the desired goal.At last, it generalized finished work in this paper and prospected the more required points of the system.